In a cellular network, there may be areas with “high traffic”, i.e. a high concentration of users. An exemplifying cell 100 comprising areas 103 with a high concentration of users is illustrated in FIG. 1a. In such high traffic areas 103 it may be desired to deploy additional capacity in order e.g. to keep the user satisfaction. Capacity could be added in the form of an additional macro base station, generating/serving a cell which covers one or more of the area(s) in need of extra capacity. Capacity could also be added in the form of additional nodes with lower output power, as compared to a macro base station, and thus covering a relatively smaller area, to which the desired capacity boost is concentrated.
There may also be areas, e.g. within a macro cell, with unfavorable radio conditions or “bad coverage”, where there may be a need for coverage extension. One way to achieve a coverage extension is to deploy an additional node, e.g. a node with a low output power, which concentrates the coverage boost to a relatively small area, e.g. where it is most needed.
One argument for choosing nodes with lower output power for increasing capacity or coverage as in the above cases is that the impact on the “original” macro nodes/network can be minimized. That is, by that the interference to an “original” macro node, with a coverage which at least partially overlaps the coverage of the “added” lower output power node, may be limited to a relatively small area.
FIG. 1b illustrates a macro base station 102, which provides a wide area coverage 100 (also called macro cell). FIG. 1b also shows examples of low power nodes that are deployed to provide small area capacity/coverage. In this example, pico base stations 104, relays 108 and home base stations 112 (femto cells) are shown, with respective coverage areas 106, 110, and 114. A pico base station can either be similar to a macro eNB, but typically with more limited coverage, for example, having a lower max transmission power, or, be a remote radio unit connected to a main unit. A common term for such pico/relay/femto cells is “underlay cells”, served by “underlay nodes”. This type of network deployments are typically referred to as: “Heterogeneous Networks”, “multilayer networks” or shortly “HetNets”.
Underlay cells typically operate at lower reference (pilot/perch) signal powers, as compared to macro cells. This means that if the cell selections as well as mobility decisions are based on received reference signal strengths, the downlink cell border will be located closer to the underlay node than to the macro node/base station/eNB. If the uplink sensitivity for all cells is similar, or if the difference in uplink sensitivity is not equivalent to the difference in reference (pilot/perch) signal powers, then the uplink cell border will be different from the downlink cell border.
FIG. 2 illustrates a scenario where the uplink 204 and downlink 206 cell borders are separated. A situation where the UL and DL borders are separated may be referred to as an uplink/downlink (or downlink/uplink) imbalance in the area between the separated borders. This means that a UE in the area between the separated borders will have a better uplink connection to the underlay node 208, but because of the stronger DL transmit power of the macro node 202, it will receive a stronger DL signal from the macro node 202. The situation of uplink/downlink imbalance is not limited to macro cell/underlay cell combinations, but may arise also between macro cells and in locations with unfavorable radio conditions, e.g. in urban environments.
Referring to the example illustrated in FIG. 2, a first UE served by the macro node 202 may cause significant uplink interference to the underlay node 204 if located in an area relatively close to the underlay node. In fact, if located in the area with uplink/downlink imbalance, said UE may even have the best uplink to the underlay node/cell, but might nonetheless not have detected the underlay cell reference signal.
One way to relieve this situation of significant interference to the underlay node is to consider an underlay cell range expansions by considering offsets in the cell selection and/or mobility decisions. Thereby, potentially interfering UEs served by the macro node will be at a longer distance away from the underlay node, and thereby induce less interference to the underlay node. However, this also means that some UEs served by the underlay node can be subjected to critical interference from the macro node in the downlink.
Concerning relays, relaying support was added in the Rel-10 version of 3GPP LTE specification. The relay solution described in Rel-10 is a so-called “layer 3 relay”, which means that all radio protocols (layers 1-3) are terminated in the Relay Node (RN). UEs connect to the RN over standard Uu interface, meaning that backwards compatibility with Rel-8 UEs is achieved. From a UE perspective, the RN looks like an ordinary eNB. The RN has no fixed backhaul, but connects wirelessly to a donor cell using the Un interface. The donor cell is controlled by a donor eNB and is based on Uu protocols, with some modifications. The donor eNB also serves UEs connected directly to the donor eNB.
So-called “inband relays” operate in the same frequencies as the macro layer (i.e. same as the donor node), which implies that the same frequency range is used on the access link and backhaul link. One issue with these relays is the uplink/downlink imbalance problem, which comes from the fact that the relay uses a lower transmit power than the macro eNBs, as described above and illustrated in FIG. 2. Because of its lower transmit power, the size of the relay cell, when measured based on downlink Reference Symbol Received Power (RSRP), is smaller than the macro cell. Still, when considering uplink transmissions, UEs connected to the macro cell and located close to the relay cell may cause interference to UEs connected to the relay cell, as previously mentioned.
One solution to the uplink/downlink imbalance problem is to base the cell selection on UE measured path loss, instead of RSRP. In that case, the relay cell size is effectively increased (cell range extension), so that all UEs that have a better uplink to the relay connect to it. Further, an offset to the RSRP measurement can be used to increase the cell size of the relay cell. Users in cell range extension will then suffer from downlink interference from the macro eNB transmitting at higher power than the relay node. This situation needs to be solved by coordinating downlink transmissions between the macro and the relay cells in either frequency or time domain. Solutions for this were introduced in Rel-10, e.g. Carrier Aggregation can be used in frequency domain, and Almost Blank Subframes in time domain.
An alternative way to build a relay and to avoid the uplink/downlink imbalance problem would be to provide relaying only in the uplink, which is illustrated in FIG. 3. FIG. 3 illustrates an uplink only relay, UORN, associated with a donor eNB. The UORN is further associated with an uplink relay area, 302, in which area UEs may be served in the uplink by the UORN and in the downlink by the eNB. Thus, the downlink would still be received from the (macro) eNB. With this solution, only uplink coverage can be extended, but because of its lower transmit power, the UE is expected to be uplink coverage limited (rather than downlink coverage limited).
Such an uplink only relay would terminate only layer 2 protocols, i.e. up to MAC layer. Such layer 2 relay nodes were discussed during early phases of the Rel-10 standardization, but were abandoned in favor of the layer 3 relay described above.